The proposed work focuses on the further development of advanced methods in nuclear magnetic resonance (NMR) spectroscopy for metabolomics-based studies. NMR-based metabolomics has been shown to be a powerful methodology for identifying metabolic perturbations in a variety of different biological states and sample types. Along with mass spectrometry, NMR spectroscopy is a robust and reproducible platform for conducting metabolite profiling for a number of applications including early disease diagnosis, treatment monitoring, drug development and basic investigations in systems biology. While the metabolome is known to provide an instantaneous snap-shot of biological status, the identification and validation of potential biomarkers of health and disease is challenging due to the complexity of their overlapping signals in biological samples. During the past grant period we developed a new set of advanced NMR tools that can be brought to bear on this problem, and showed that we can significantly increase the ability of NMR to dissect the complex samples and identify sensitive and specific metabolite biomarker candidates. In particular we showed that by using an isotope tagging strategy we could improve the limit of detection by 5 to 10-fold and increase the number of NMR-detectible metabolites by a similar factor. We showed that the NMR and mass spectrometry can be used together in a complimentary fashion to build highly accurate metabolite profiles. We also showed that selective TOCSY could be used to eliminate uninteresting background signals to improve biomarker discovery and the classification of samples. And we developed LC-NMR and microcoil NMR methods for metabolite identification. The current proposal focuses on the bringing these approaches together to build out a robust NMR platform for quantifying over 100 metabolites in a blood, urine or cell extract sample. We will develop a comprehensive approach for unknown identification including an expanded library of isotope tagged metabolites. And we propose to miniaturize this system such that it can be run on a modest amount (100 L) of sample while maintaining an ability to quantify approximately 50 metabolites. The full development of this approach, as proposed here, would change the current paradigm in NMR-based metabolomics and provide an even stronger complement to current MS-based methods. Validation of these methods on a set of commercial serum samples, urine and cell extracts, and a comparison of these new approaches to current methods are also discussed. If this work is successful, we will have provided a much improved NMR platform for advanced metabolomics studies that will be applicable to a range of studies from early disease detection and therapy monitoring to basic studies of systems biology. PUBLIC HEALTH RELEVANCE: This project seeks to develop a reliable and advanced platform for identifying biomarkers of disease in serum and urine samples and for fundamental studies of biological systems. When fully developed these approaches could be used to detect small molecules in the blood or urine that indicate the onset of various diseases.